
Explore high availability in Azure, from defining availability and SLAs to a deep dive into Azure high-availability services with practical demos and a design lifecycle.
Explore the concept of availability, uptime, and how percent availability translates into minutes of downtime, why 100% uptime is unattainable, and how three nines to five nines demand automated recovery.
High availability eliminates single points of failure and ensures continuous operations, requiring active, custom design of your architecture beyond Azure's building blocks, and recognizing the weakest link.
Learn how Azure service level agreements guarantee availability, how credits apply when targets fall short, and how compound SLAs and redundancy with load balancing boost overall uptime.
Discover how Azure VM service level agreements depend on disk choices, from standard HDD to premium SSD, affecting uptime from 95% to 99.9% and cost.
Learn how Azure VM SLAs depend on storage type and deployment topology, comparing availability sets and availability zones to achieve uptime from 99.95% to 99.99% with multiple instances.
Explore Azure sql database availability across basic, standard, and premium tiers, including zones and geo replication. Learn how sla details and configuration choices shape financial guarantees and high availability.
Explore Azure high-availability components, including load balancing, failover clusters, zone deployment options, availability zones, and deployments in regions to design resilient apps with automation and no single point of failure.
Explore how public and internal load balancers reduce single points of failure and how cross-region designs with app services, vm scale sets, Windows failover clusters, and application gateway boost availability.
Define and document your high-availability requirements during the design stage, considering planned downtimes, cost and complexity tradeoffs, multi-region replication, and clear availability metrics.
Design your high-availability solution by applying patterns from the Azure Architecture Center, avoiding single points of failure, and planning for failover, caching, and automation.
Test high availability with load testing to reveal bottlenecks, validate backups and failover readiness, and use Chaos Studio to simulate outages while validating monitoring and automation.
Frequent automated deployments and thorough backups improve reliability; detect deployment failures via pipelines, identify increased production errors, and roll back faulty releases to minimize Azure outages.
Detect production failures with independent health probes behind load balancers, and use Azure Monitor with Log Analytics to trigger alerts and dashboards, applying a selective diagnostics strategy.
Explore high availability concepts in Azure from a Portugal user group session. Share feedback via course Q&A or Microsoft Azure unofficial user group on Facebook to help improve future content.
In late 2022, I was asked to present to a local Azure User Group in Portugal. Unfortunately, we were not allowed to video record the talk to release it for people to view after the event. If you did not see it in person, or live online at the time, then you would have missed it.
So I decided to record it myself from home and post it here at Udemy.
Some important caveats before you enroll in this course:
I created this presentation for an in-person event. If I was making it as a course, I might have structured things a bit differently. But ultimately, this is a presentation with slides and talking.
Unlike my other Azure courses, this course does not prepare you for any official Microsoft certification or exam. None of the content aligns with any official Microsoft exam.
And given only an hour or so to present, I could not get into deep technical details like building a solution inside Azure itself. This is intended to be an overview to inspire further research.
The course covers four overall topics:
The concepts of availability and high availability (HA)
The concept of Service Level Agreement (SLA)
High availability services in Azure
Five Steps to a High-Availability Solution
You should be able to watch the course all at once, and you are not expected to have an Azure account or follow-along with any hands-on demos.